C++ network sockets, SCTP and packet size -

-

i'm developing server using connection-oriented sctp serve small number of clients. after finishing first prototype naive implementation, i'm profiling application optimize. turns out, 1 of 2 main consumers of cpu time networking part.

there 2 questions efficiency of application-level protocol have implemented:

1) packet size

currently, use maximum packet size of 64 bytes. can find many posts discussing packet sizes big, can small? sctp allows me read 1 packet @ time - upd - while guaranteeing in-order delivery - tcp - simplified implementation significantly. however, if understand correctly, cost 1 syscall each , every time send packet. amount of syscalls have significant impact on performance? able shave off lot of cpu cycles sending messages in bunches in bigger packets, i.e. 1024 - 8192 bytes?

2) reading , writing buffers

i'm using memcpy move data , out of application-level network buffers. found many conflicting posts more efficient, memcpy or normal assignment. i'm wondering if 1 approach faster other in scenario:

option 1

void network::receivepacket(char* packet) {     uint8_t param1;     uint16_t param2     uint32_t param3;      memcpy(&param1, packet, 1);     memcpy(&param2, packet+1, 2);     memcpy(&param3, packet+3, 4);      // handle packet here }  void network::sendpacket(uint8_t param1, uint16_t param2, uint32_t param3) {     char packet[7]      memcpy(&packet, &param1, 1);     memcpy(&packet+1, &param2, 2);     memcpy(&packet+3, &param3, 4);      // send packet here }

option 2

void network::receivepacket(char* packet) {     uint8_t param1;     uint16_t param2     uint32_t param3;      param1 = *((uint8_t*)packet);     param2 = *((uint16_t*)packet+1);     param3 = *((uint32_t*)packet+3);      // handle packet here }  void network::sendpacket(uint8_t param1, uint16_t param2, uint32_t param3) {     char packet[7]      *((uint8_t*)packet) = param1;     *((uint16_t*)packet+1) = param2;     *((uint32_t*)packet+3) = param3;      // send packet here }

the first 1 seems lot cleaner me, i've found many posts indicating maybe second 1 quite bit faster.

any kind of feedback of course welcome.

as far know compilers optimize memcpy calls in particular should use it.

about first question:

  • summary: make packet size big can , avoid possibility of having enlowered cpu performance.

a syscall, system call, os replying or processing request , every time request being executed in kernel, moderate amount of work. honest not familiar sctp concept, matter of fact haven't dealt socket programming since last time worked on stuff , created server via tcp. remember mtu relevant physical layer element 1500, recall implementing packet size 1450-1460, trying maximum packet size underneath 1500 cap.

so im saying if want os less active wont trouble cpu performance.


Comments

Post a Comment

Popular posts from this blog

java - activate/deactivate sonar maven plugin by profile? -

-
i use sonar plugin executes if activated in maven profile. sonar runs on every compile. the problem build fail other team members not have sonar application installed locally. pom.xml checked in team members same copy including sonar configuration, don't want them modify purpose of disabling sonar (they might inadvertently check modified copy in). please advise how configure maven (.m2/settings.xml) and/or pom.xml sonar feature can enabled/disabled local profile. here sonar configuration in each project's pom.xml: <plugin> <groupid>org.codehaus.mojo</groupid> <artifactid>sonar-maven-plugin</artifactid> <version>1.0</version> <executions> <execution> <id>make-sonar</id> <phase>compile</phase> <goals> <goal>sonar</goal> </goals> </execution> </executions> ...
Read more

python - TypeError: can only concatenate tuple (not "float") to tuple -

-
i'm new python, question might easy, anyway, patience: as trying call newton-raphson method calculate implied volatility in black-scholes formula call/put option pricing, first thing is, newton method in scipy.optimize seems calculate function's zeros, in black-scholes formula, want function's value option price, not zero. (i'm new here programming, i'm not sure techniques.) should write function things like: def f(sigma, price): return bsformula(s0,k,r,t,q,sigma) - price then, while calling newton method, takes args=() 1 parameter in function, write this: value = newton(bsprice2, 0.5, args=price) but error message: file "bs.py", line 36, in bsimpvol value = newton(bsprice2, 0.5, args=float(price)) file "/usr/lib/python2.7/dist-packages/scipy/optimize/zeros.py", line 143, in newton q0 = func(*((p0,) + args)) typeerror: can concatenate tuple (not "float") tuple could tell me why that? how fix it? apprecia...
Read more

java - What is the difference between String. and String.this. ? -

-
this question has answer here: meaning of .this , .class in java 2 answers i'll straight point. im still learning bit of syntax , want know difference between code is code a: public class buttonz extends jbutton{ public buttonz(){ settext(new string(string.valueof(i))); } } please ignore fact i undeclared, not lost. code b: public class buttonz extends jbutton{ public buttonz(){ settext(new string(string.this.charat(i))); } } what don't yet understand difference in typing string.this , string. i under assumption when use dot operator on class accessing it's static methods(and/or variables if they're not hidden). i have studied little bit , have concluded when using string. accessing string static methods.. when using string.this. i'm accessing methods class buttonz extendi...
Read more